Research on Olympic Games Hosting Strategy Based on Machine Learning Algorithm

With the increasing size and cost of the Olympic Games and the declining willingness and enthusiasm of many cities to bid for the Games, the IOC is facing a decline in the number of bids for the Summer and Winter Olympic Games. In order to explore an innovative incentive for countries and cities to bid for the Olympic Games, we developed an Olympic Games weighting model, and an Olympic Games permanent venue evaluation clustering model. First, to construct the Olympic Games impact evaluation model, we collected data on 13 indicators in 7 dimensions from 18 Olympic Games host cities, calculated the impact coefficients of each dimension through gray prediction and difference calculation, and summarized the impact of the Olympic Games on host cities through a coupled model based on AHP-EWM(AE):Hosting the Olympic Games has the greatest impact on economy, reputation, and urban development opportunities, and The impact on the environment is the least. Second, in order to select the permanent host of the Olympic Games and to investigate possible changes, we calculated the scores of cities that have already hosted the Olympic Games using the TOPSIS method and predicted the scores of non-bidding countries using the Lasso regression algorithm, collecting data on 13 indicators for the next 18 Olympic bidding countries. Then, we clustered and ranked the non-bidding countries with the bidding countries using a K-means clustering model to select the location of the permanent Olympic host country; finally, we used AE to calculate the weights of the cities most suitable as permanent Olympic host countries for comparison and found that this strategy would significantly boost the tourism aspect of the host city (14.80%), but also hinder its potential (-16.31%).